Are Nuclear Rockets "Cheating"?

[Cocox]

They're based in reality and are balanced by having a low TWR and high cost. So no.

Depends.. I edited the tech tree and made it insanely expensive to research.. no more 1 mission to minmus and the ultimate answer is received..

Also edited some engines to 1500 ISP, but on ironman mode and at 1.5m blingbling each.. it's fun and little room for mistakes..

[juanml82]

I prefer not to use LV-N's, not because I think it's "cheating" but because they make it a bit too easy for me (and just me; y'all have fun however you like). I'd probably change on that if they gave the LV-N its own fuel type and attendant hassles.

Well, the fuel is liquid hidrogen, it's just that they don't use oxidizer.

Which means they'd be even more powerful, as almost half of the fuel tank is oxidizer in KSP (unlike RL).

I think the game balance is only 1.5m parts, can't mount them in a decoupler with stuff below them or the fairings will hit each other during staging, risk of the engine getting destroyed by 2.5 or 3.5 m lower stages if the TWR is too big and long burn times.

[davidpsummers]

Thanks for the info from everyone! I had no idea...

[Jarin]

If any stock part is cheating due to being unrealistic it would be jets. Real jets have similar Isps but don't classify intakeair as reaction mass. KSP jets do, and thus are much more efficient.

Wait... isn't that the entire point of atmospheric engines? Not needing reaction mass, because you're just taking in air and pushing it backwards?

[Levelord]

Wait... isn't that the entire point of atmospheric engines? Not needing reaction mass, because you're just taking in air and pushing it backwards?

I just had a facepalm moment when I read No one's post too. I think he's forgetting that the oxidizer is simply liquid oxygen that is otherwise freely available in the atmosphere and jets make use of this, as they should.

[cantab]

IIRC it's more a definition thing. For KSP's jets the specific impulse is set considering the fuel consumption, but then applied considering the fuel+air consumption. It effectively saves the fuel twice over.

[ferram4]

On the jet thing...

Well, considering that the jets use 15 kg of air for every 1 kg of LF, that means that they effectively have an Isp 16 times greater than what is reported. This is because in a proper analysis of a jet engine, when considering mass flow you would have the mass flow of air out balanced by the mass flow of air in through the intake, resulting in the only change in mass of the vehicle being the fuel flow, so it's the only significant bit of mass flow to consider.

A good way to show yourself that this is the only way that makes it kosher is that if you count the entire mass flow out the exhaust, then exhaust velocity = Isp * g0. Now, if you have a jet that has an Isp of 4000s (approximately correct for a low-bypass turbofan), if the entire mass flow were counted towards the exhaust velocity, this would imply that it was shooting air out the back at 4000s * 9.81 m/s² = 39,240 m/s. Which is 3.5 times Earth's escape velocity. Obviously, that interpretation makes no sense, so logically, we can determine that the standard definition of Isp for jets must use is mass flow of fuel alone, not total mass through the engine.

But in KSP, they use total mass through the engine. You can check this yourself by going and changing the IntakeAir requirements of a jet part and see how much fuel it consumes if it's demanding 15 air per unit LF or 2 air per unit LF.

TL;DR: Current stock jets have "true" exhaust velocities greater than Kerbin escape velocity, and at peak performance, greater than Earth escape velocity (11.2 km/s). Somehow reacting less mass they produce more energy to accelerate remass than rocket engines do, and this is broken.

On the original NTR-being-balanced-or-bot discussion, It is heavy, and its TWR is terrible compared to the real-life concept. On the other hand, it can use storable propellants instead of LH2. Once you consider how OP LF / O is as a propellant combination (good density, perfectly storable, peak Isp of 390s in non-NTRs), the performance of the stock NTR is actually pretty-well on par. I'd prefer needing special propellant for it and giving it a bit of a serious buff in terms of thrust or Isp to make up for that, but I find it odd to run an NTR on the same stuff as a standard motor.

[Renegrade]

This. The idea of a NERVA engine that uses oxidizer is ... stupid, to say the least. Even using monoprop would be more sensible :/

There's been a couple of posts like this throughout the thread.

I don't see why an NTR couldn't "burn" oxidizer. From my understanding, LH2 is usually chosen for NTR applications since it has a low molecular weight (helps it attain high exhaust velocity), but in theory you could just heat up any old gas or fluid and make thrust. In the simplest terms, an NTR is like the jet of steam that comes out of a boiling kettle. A very hot, nuclear kettle. With extremely hot, fast steam.

I imagine that building an NTR with extremely corrosion-resistant materials might oh say, make it heavier and limit it's thrust somewhat though *cough,cough*.

[r_rolo1]

There's been a couple of posts like this throughout the thread.

I don't see why an NTR couldn't "burn" oxidizer. From my understanding, LH2 is usually chosen for NTR applications since it has a low molecular weight (helps it attain high exhaust velocity), but in theory you could just heat up any old gas or fluid and make thrust. In the simplest terms, an NTR is like the jet of steam that comes out of a boiling kettle. A very hot, nuclear kettle. With extremely hot, fast steam.

I imagine that building an NTR with extremely corrosion-resistant materials might oh say, make it heavier and limit it's thrust somewhat though *cough,cough*.

Yeah, true, there is nothing by definition that forbids using a oxidixer agent ( most likely liquid oxygen ) as propelling agent in a NERVA. But the efficiency of a NERVA is a inverse function of the molecular mass of the propelant ( better said, it is proportional to the average speed of the molecules of the propeller, that for a certain fixed temperature is a inverse function of the square root of the molecular mass of it ( IIRC ) ) , thus using LO2 is probably more than a order of magnitude worse than using LH2 ( depends of exaust temp and some other stuff ) and it is even worse with other ( heavier ) oxidizing agents. But besides that, it is really a bad idea to put anything in the proximity of a gaseous superheated oxidizer ... we know that because we use that exact same process to purge impurities from steel and to in general burn stuff to not even ashes:

Now think on that inside your NERVA ( not only on the outside ) ... remember that your typical NERVA will need to have a large heat condutivity ( otherwise the nuke reactor would not be able to heat the propeller ), that normally means metal alloys that react pretty well with oxygen at high temps ( well, there is always aluminium nitride, but even that will react with oxygen above 1400 ºC ), or worse in the models that have propeller gas passing direcly on the core as moderator/cooler ( uranium and plutonium burn quite well ... ). In other words, if you use a oxidizer inside a NERVA you will most likely burn it to uselessness ...

Edited August 24, 2014 by r_rolo1

[cantab]

The NERVA needs a light molecule for its propellant to get the high Isp. The temperature of the exhaust is actually lower than a chemical rocket.

But the current situation is to save Squad the work of making a new resource and new tanks just for one engine.

I wouldn't mind seeing a "Cryogenic fuel" that's used alone in the LV-N and with ox in some other high Isp chemical engines, and that slowly boils off. It would change the gameplay quite a bit though - LV-N's would be good for kerbin departure stages but no longer suitable for orbital insertion at the target.

[No one]

Wait... isn't that the entire point of atmospheric engines? Not needing reaction mass, because you're just taking in air and pushing it backwards?

I just had a facepalm moment when I read No one's post too. I think he's forgetting that the oxidizer is simply liquid oxygen that is otherwise freely available in the atmosphere and jets make use of this, as they should.

From Wikipedia's page on Specific Impulse:

When calculating specific impulse, only propellant that is carried with the vehicle before use is counted. For a chemical rocket the propellant mass therefore would include both fuel and oxidizer; for air-breathing engines only the mass of the fuel is counted, not the mass of air passing through the engine.

The intake air is how they get their additional Isp. Did you think that for some reason jets just magically had Isps of insanely high for no reason? Also, if it did work that way, then why has NASA not taken a fuel tank full of air and used them in space? In KSP we don't have fuel tanks full of air but if we did then jet engines would be better than nuclear rockets.

[r_rolo1]

The NERVA needs a light molecule for its propellant to get the high Isp. The temperature of the exhaust is actually lower than a chemical rocket.

But the current situation is to save Squad the work of making a new resource and new tanks just for one engine.

I wouldn't mind seeing a "Cryogenic fuel" that's used alone in the LV-N and with ox in some other high Isp chemical engines, and that slowly boils off. It would change the gameplay quite a bit though - LV-N's would be good for kerbin departure stages but no longer suitable for orbital insertion at the target.

The exaust temperature of the ingame LV-N is surely lower than the chemical engines exaust, but it is hot enough inside the reactor to make the exterior red hot ( ~500-600 ºC ) and the exaust itself should be certainly above 1000 ºC. In those situations using oxygen as propeller would basically make the same effect as igniting a blowtorch inside the reactor. But even in far lower temperature ( like the real one of NERVA ) it would be highly unwise to pass pure oxygen through metal surfaces ... and OFC, it would be stupid to use oxygen when hydrogen is avaliable

On the Cryo fuel suggestion, tha is a good idea, but it would definitely need some work on the balance ... I would really like it, though.

[sedativechunk]

No, nuclear rockets are NOT cheating. They are a stock (vanilla) part of KSP and they actually have been built and tested in ground firings in real life.

The nuclear rocket is a real concept when talking about interplanetary travel. Back in the Apollo days, they were building and testing actual nuclear rockets in the American desert. They are truly efficient rockets and one was even proposed for the Apollo rocket for traveling to the moon as the upper stage portion. Unfortunately to this date, a nuclear rocket has never actually been launched in space in fear of catastrophe. If the rocket were to explode while in Earth's atmosphere, there is a risk of radioactive waste covering a widespread area of land.

Look in amazon video if you have prime, there are some interesting documentaries on the nuclear rocket program in the 60's. The facility where they tested these rockets still exist today, untouched all these years. If you want to talk about "cheating", the one mod I consider cheating to an extent is the alcubierre/warp drive. Wrap drives are theoretically possible but I sincerely doubt humanity will ever actually build one, at least not in any of our lifetimes.

[Masetto]

It's really not that imba.

I avoid using them a lot of the time, so I can have quicker burns (using the Poodle typically, or maybe the 909).

The loss in DeltaV is not so substantial. It's still plenty easy to get wherever you need to go without them, and often the rocket looks better as well.

[Renegrade]

Yeah, true, there is nothing by definition that forbids using a oxidixer agent ( most likely liquid oxygen ) as propelling agent in a NERVA. But the efficiency of a NERVA is a inverse function of the molecular mass of the propelant ( better said, it is proportional to the average speed of the molecules of the propeller, that for a certain fixed temperature is a inverse function of the square root of the molecular mass of it ( IIRC ) ) ,

Yeah, that's what I meant by 'LH2 / lower molecular weight', and I know that an oxidizer-burner would be rather risky. I believe Russian turbopumps tend to run oxidizer-rich so I imagine there are engineering techniques that can mitigate some or all of the issues....Also don't forget that Kerbals engineer with some crazy substances.. 3600c max temp? Some sort of non-reactive super-carbon?

By the way, this talk of burning oxidizer reminds me of this screwy idea I had: The developers have stated (jokingly) that Eve's oceans are LiquidFuel. So I had a thought - does that mean that Eve's atmosphere is liquid fuel vapor? Could you then make a jet engine that works on Eve that intakes fuel instead of air and mixes it with stored oxidizer to produce thrust?

The NERVA needs a light molecule for its propellant to get the high Isp. The temperature of the exhaust is actually lower than a chemical rocket.

It's lower temp? Really? That I didn't know.

But the current situation is to save Squad the work of making a new resource and new tanks just for one engine.

Unfortunately quite true, and I think that could be said about a great many things .. :/

If they wanted it closer with minimum effort though, they could actually just make it run off of LiquidFuel only. I think removing the Oxidizer stanza and changing the LF ratio to '1.0' would let it run, no? The mass fraction of the tanks would well, excuse the pun, tank, but we can Tweak away oxidizer since uhh.. 0.23? 0.23.5?, so it's not like it has to carry a bunch of useless Oxidizer around with it anymore..

Maybe I'll try modifying it in game to see if that works, heh

I wouldn't mind seeing a "Cryogenic fuel" that's used alone in the LV-N and with ox in some other high Isp chemical engines, and that slowly boils off. It would change the gameplay quite a bit though - LV-N's would be good for kerbin departure stages but no longer suitable for orbital insertion at the target.

I think that would be neat. I played with RealFuels for a while and I liked how it had that distinction. Unfortunately I mostly play 1/10th scale and it's oriented towards full-scale more, and I found the huge list of fuels a bit overwhelming, but I wouldn't mind seeing some fuel distinctions in KSP. Having a "CryogenicFuel" to go with "LiquidFuel" and "SolidFuel" would sorta match up in terms of KSP design, plus it matches up (kinda, if you consider 'cryo' to be a 'gas' hehe) with the classic states of matter.

Another thing though - from my chemistry classes all so long ago, I seem to recall that you can keep a gas in a liquid state if it's under enough pressure.. How heavy would a pressure vessel have to be in order to keep LH2 liquid?

[SorryDave]

The use of Nuclear Thermal Rockets has been a key part of many versions of the Mars reference design by NASA.

I imagine that the KSP version burns both an oxidizer and a fuel before super heating the exhaust with a nuclear source.

[Renegrade]

The loss in DeltaV is not so substantial. It's still plenty easy to get wherever you need to go without them, and often the rocket looks better as well.

This here above is a good point.

If you swapped out three or four LV-Ns (180-240kn) to a single Poodle (220kn), you'd lose about 50% DV due to the (slightly more than) halving of the specific impulse, but you could gain a fair amount of that back in terms of mass fraction (a poodle is less mass than a single LV-N now, having two of three or more LV-Ns is adding a lot of dry mass).

Swapping out a single LV-N for like a single poodle would have roughly the same mass fraction, but the Poodle ship could leverage the Oberth effect better.

It's not quite as cut and dried as people might think

[r_rolo1]

That was always one of my main gripes about the people thoughts about the LV-N ( that is , the solution to all problems as long as you are in orbit ) : the weight of the engine and it's low thrust make it unsuitable in various niches, like small ships or, like you said, situations where you would actually prefer a higher thrust. 2.25 mass is a lot, more than a LV-909+ FT L 400 + Mk1 lander can ( that is enough to get to Duna orbit from the surface and come fack to Kerbin if you navigate well enough ) and that is even worse if you make the bad call of spamming LV-N ( yesterday I've seen a video of someone making a Eve and back mission ... his transfer stage had 26 (!) LV-N ) to get the thrust you desire ...

[Red Iron Crown]

This here above is a good point.

If you swapped out three or four LV-Ns (180-240kn) to a single Poodle (220kn), you'd lose about 50% DV due to the (slightly more than) halving of the specific impulse, but you could gain a fair amount of that back in terms of mass fraction (a poodle is less mass than a single LV-N now, having two of three or more LV-Ns is adding a lot of dry mass).

Swapping out a single LV-N for like a single poodle would have roughly the same mass fraction, but the Poodle ship could leverage the Oberth effect better.

It's not quite as cut and dried as people might think

It's also not either/or. Put four LV-Ns around a poodle and you get a cluster with over 600s of Isp and twice the TWR of an LV-N-only cluster, for example; or two LV-Ns and an aerospike has about 550s and even better TWR. Heterogeneous clusters let you tune the balance of Isp and TWR to your liking.

[Ravyne]

Honestly, you could qualify anything at all as 'cheating'.

Reaction wheels make it easier to maneuver!

Solar panels let you regenerate electrical power!

Engines let you go fast!

So really, this is moot.

[Renegrade]

and that is even worse if you make the bad call of spamming LV-N ( yesterday I've seen a video of someone making a Eve and back mission ... his transfer stage had 26 (!) LV-N ) to get the thrust you desire ...

Twenty six?? Replace those a single mainsail, and he's actually got more* delta-v, and more acceleration, since he just lost like 50 tons of engine!

What was he sending to Eve? The CN Tower? :S

* in most situations that might ever reasonably occur. I'd whip out the old abacus and figure out what those situations are, but I'm overcome with 'meh' about that..

It's also not either/or. Put four LV-Ns around a poodle and you get a cluster with over 600s of Isp and twice the TWR of an LV-N-only cluster, for example; or two LV-Ns and an aerospike has about 550s and even better TWR. Heterogeneous clusters let you tune the balance of Isp and TWR to your liking.

Yes, this is a good point too. People have to drop the 'spam LV-N' stuff if they want effective rockets..

[r_rolo1]

Twenty six?? Replace those a single mainsail, and he's actually got more* delta-v, and more acceleration, since he just lost like 50 tons of engine!

What was he sending to Eve? The CN Tower? :S

Almost Someone posted in the Spacecraft Exchange asking if anyone had the craft file of the ship in this video ( note that is definitely pre 23.5 , but still ... and BTW the image in the still is the transfer stage ):

• 
  

For curiosity, I've seen the whole thing ( because my philosophy about making rockets is the exact oposite of those monster rockets ) ... Holy Lagfest at first, transfer stage with 26 LV-N ( mainly because the guy wanted more thrust, like I was mentioning before ) ... God almighty, some people really think "MOAR boosters"

Edited August 24, 2014 by r_rolo1

[Red Iron Crown]

I did a 24 LV-N design for a big single-stage mothership once. The ship was about 600t and the dV was about 9km/s, TWR about 0.25. "Moar LV-Ns!" is sometimes a good strategy, but it is certainly true that it isn't always the best.

[Renegrade]

For curiosity, I've seen the whole thing ( because my philosophy about making rockets is the exact oposite of those monster rockets ) ... Holy Lagfest at first, transfer stage with 26 LV-N ( mainly because the guy wanted more thrust, like I was mentioning before ) ... God almighty, some people really think "MOAR boosters"

That is most certainly moar boosters there hehe. I'll have to watch that later, and maybe see if I can design something a bit less crazy and 40-million-funds-y.

I did a 24 LV-N design for a big single-stage mothership once. The ship was about 600t and the dV was about 9km/s, TWR about 0.25. "Moar LV-Ns!" is sometimes a good strategy, but it is certainly true that it isn't always the best.

Ouch.. Wait, that would have to be in orbit, as it wouldn't lift off with only 0.25TWR.. so uhh.. what's in the 600T? :S

I have some 2400T units, but those are surface lifters laboring under heavy BTSM upper stages (which are .. upwards of 200T or so, mostly life support and heavy injection burn units)... not even close to 600T in orbit though.. My usual stock units are ah, probably not much more than 40T-ish max (those are the mobile science lab/refuel stations).

[Red Iron Crown]

Ouch.. Wait, that would have to be in orbit, as it wouldn't lift off with only 0.25TWR.. so uhh.. what's in the 600T? :S

It was a Jool-5 attempt (that failed for reasons other than not enough dv), it carried a modular lander system plus a lab and hab space.

In orbit, moar nukes visible:

The lifter that put it there (~3600t, 16 SLS main engines):